Atmos. Meas. Tech., 6, 1217–1226, 2013www.atmos-meas-tech.net/6/1217/2013/doi:10.5194/amt-6-1217-2013© Author(s) 2013. CC Attribution 3.0 License.

EGU Journal Logos (RGB)

Advances in Geosciences

Open A

ccess

Natural Hazards and Earth System

Sciences

Open A

ccess

Annales Geophysicae

Open A

ccess

Nonlinear Processes in Geophysics

Open A

ccess

Atmospheric Chemistry

and Physics

Open A

ccess

Atmospheric Chemistry

and Physics

Open A

ccess

Discussions

Atmospheric Measurement

TechniquesO

pen Access

Atmospheric Measurement

Techniques

Open A

ccess

Discussions

Biogeosciences

Open A

ccess

Open A

ccess

BiogeosciencesDiscussions

Climate of the Past

Open A

ccess

Open A

ccess

Climate of the Past

Discussions

Earth System Dynamics

Open A

ccess

Open A

ccess

Earth System Dynamics

Discussions

GeoscientificInstrumentation

Methods andData Systems

Open A

ccess

GeoscientificInstrumentation

Methods andData Systems

Open A

ccess

Discussions

GeoscientificModel Development

Open A

ccess

Open A

ccess

GeoscientificModel Development

Discussions

Hydrology and Earth System

Sciences

Open A

ccess

Hydrology and Earth System

Sciences

Open A

ccess

Discussions

Ocean Science

Open A

ccess

Open A

ccess

Ocean ScienceDiscussions

Solid Earth

Open A

ccess

Open A

ccess

Solid EarthDiscussions

The Cryosphere

Open A

ccess

Open A

ccess

The CryosphereDiscussions

Natural Hazards and Earth System

Sciences

Open A

ccess

Discussions

Design and performance of a Nafion dryer for continuous operationat CO2 and CH4 air monitoring sites

L. R. Welp1, R. F. Keeling1, R. F. Weiss1, W. Paplawsky1, and S. Heckman2

1Scripps Institution of Oceanography, University of California San Diego, 9500 Gilman Dr., La Jolla, CA, USA2Earth Networks, Inc., 12410 Milestone Center Drive, Germantown, MD, USA

Correspondence to:L. R. Welp (lwelp@ucsd.edu)

Received: 30 June 2012 – Published in Atmos. Meas. Tech. Discuss.: 7 August 2012Revised: 5 April 2013 – Accepted: 18 April 2013 – Published: 14 May 2013

Abstract. In preparation for routine deployment in a net-work of greenhouse gas monitoring stations, we have de-signed and tested a simple method for drying ambient airto near or below 0.2 % (2000 ppm) mole fraction H2O usinga Nafion dryer. The inlet system was designed for use withcavity ring-down spectrometer (CRDS) analyzers such as thePicarro model G2301 that measure H2O in addition to theirprincipal analytes, in this case CO2 and CH4. These analyz-ers report dry-gas mixing ratios without drying the sampleby measuring H2O mixing ratio at the same frequency as themain analytes, and then correcting for the dilution and peakbroadening effects of H2O on the mixing ratios of the otheranalytes measured in moist air. However, it is difficult to ac-curately validate the water vapor correction in the field. Bysubstantially lowering the amount of H2O in the sample, un-certainties in the applied water vapor corrections can be re-duced by an order of magnitude or more, thus eliminating theneed to determine instrument-specific water vapor correctioncoefficients and to verify the stability over time. Our Nafiondrying inlet system takes advantage of the extra capacity ofthe analyzer pump to redirect 30 % of the dry gas exiting theNafion to the outer shell side of the dryer and has no consum-ables. We tested the Nafion dryer against a cryotrap (−97◦C)method for removing H2O and found that in wet-air tests, theNafion reduces the CO2 dry-gas mixing ratios of the samplegas by as much as 0.1± 0.01 ppm due to leakage across themembrane. The effect on CH4 was smaller and varied within± 0.2 ppb, with an approximate uncertainty of 0.1 ppb. TheNafion-induced CO2 bias is partially offset by sending thedry reference gases through the Nafion dryer as well. Theresidual bias due to the impact of moisture differences be-tween sample and reference gas on the permeation through

the Nafion was approximately−0.05 ppm for CO2 and var-ied within± 0.2 ppb for CH4. The uncertainty of this partialdrying method is within the WMO compatibility guidelinesfor the Northern Hemisphere, 0.1 ppm for CO2 and 2 ppb forCH4, and is comparable to experimentally determining wa-ter vapor corrections for each instrument but less subject toconcerns of possible drift in these corrections.

1 Introduction

There is increasing interest in regional greenhouse gas emis-sions estimates as stakeholders aim to reduce and verifyemissions at international, national, state and city levels(NRC, 2010). Two of the most important greenhouse gasesof interest are CO2 and CH4. Atmospheric inversion methodsprovide a means of inferring emission rates based on atmo-spheric concentration measurements, but their usefulness atthe regional level has been hampered by sparse greenhousegas monitoring locations (Butler et al., 2010; Gurney et al.,2002) and uncertainty in atmospheric transport (Houwelinget al., 2010; Lin et al., 2006; Stephens et al., 2007). Re-cently, Earth Networks, Inc. has proposed to greatly increasethe density of atmospheric surface measurements by deploy-ing a network of close to 50 continuous observation sta-tions across the United States (http://www.earthnetworks.com/OurNetworks/GreenhouseGasNetwork.aspx). It is im-portant that the data collected by this network, and others,be of high quality and meet or exceed the WMO compati-bility goals of 0.1 ppm CO2 in the Northern Hemisphere and0.05 ppm CO2 in the Southern Hemisphere, and 2 ppb CH4(WMO, 2012).

Published by Copernicus Publications on behalf of the European Geosciences Union.

1218 L. R. Welp et al.: Design and performance of a Nafion dryer

Until recently, most high-accuracy continuous CO2 mea-surements were made using non-dispersive infrared (NDIR)spectroscopic analyzers (e.g. Bakwin et al., 1998). These an-alyzers require frequent calibration and complete drying ofthe air prior to analysis. A newer approach using wavelength-scanned cavity ring-down spectroscopy (CRDS) has greaterstability, reducing the frequency of calibration, and has thepotential to eliminate the need for drying (Crosson, 2008).Water vapor interferes with CO2 and CH4 concentrationmeasurements by diluting the mixing ratios in air and bybroadening the spectroscopic absorption lines of other gases.The approach that is implemented in CRDS is to concur-rently measure H2O of the sample and use experimentallyderived water vapor correction algorithms to correct for thedilution and broadening effects on CO2 and CH4 (Crosson,2008; Rella, 2010; Baer et al., 2002).

It is difficult to accurately calibrate the absolute H2O mea-surements on any instrument due to a lack of precise ref-erence humidity generation and delivery. The CRDS instru-ments have an absolute H2O uncertainty of∼ 1 % (Chen etal., 2010) and this introduces a source of error into using asingle water vapor correction algorithm for all instruments.However, the CRDS instruments can make very precise mea-surements of relative H2O differences, so if a set of water va-por correction coefficients is determined for each instrument,the uncertainty in the CO2 and CH4 dry mixing ratios can bereduced. There have been a few studies published testing thetransferability and stability of the water vapor correction al-gorithm for Picarro CRDS instruments (Chen et al., 2010;Winderlich et al., 2010; Richardson et al., 2012; Rella et al.,2013). Chen et al. (2010) suggests that applying the same setof coefficients to multiple instruments can yield high-qualitydata even with small differences in the H2O calibration frominstrument to instrument. They found the residual error afterusing a single set of water vapor correction coefficients onmultiple instruments is below 0.05 ppm for CO2 and below0.5 ppb for CH4. Winderlich et al. (2010) conclude that thewater vapor correction for an individual instrument is stableover a year and half and estimates the repeatability of the cor-rected measurements is within 0.03 ppm for CO2 and 0.3 ppbfor CH4. Even so, the manufacturer conservatively recom-mends experimentally determining the water vapor correc-tion coefficients through repeated testing for each individualinstrument at the start of operations in order to meet WMOcompatibility goals at water levels greater than 1 % (Rella etal., 2013).

These types of water vapor correction experiments are in-herently difficult to perform because H2O and CO2 can in-teract with tubing walls in most experimental setups. CO2can adsorb to the tubing walls, and increasing H2O will dis-place that CO2, creating an artifact in dry/wet-air compar-isons used to calculate correction coefficients, especially insetups where the H2O level is changing continually. Also,the accuracy of the water vapor corrections are based on CO2and CH4 measurements of and are statistically limited by the

accuracies of the CO2 and CH4, in addition to H2O measure-ments themselves (Rella et al., 2013). Reducing the H2O inthe sample would reduce the uncertainty in the corrected dry-gas mixing ratios. The simple drying technique we presenthere does not eliminate the water vapor influence, but re-duces the water vapor correction by an order of magnitudeor more, thus eliminating the need to characterize the watervapor correction on each instrument before deployment andto monitor that calibration over time.

The Nafion membrane is known to be semi-permeableto water vapor and relatively impermeable to other gases(Leckrone and Hayes, 1997). Nafion dryers are built witha tubing of semi-permeable membrane separating an inter-nal sample gas stream from a counterflow purge gas streamcontained within a stainless steel outer shell. If the partialpressure of water vapor is lower in the purge gas stream,then water is removed from the sample gas stream. There aremany different ways to supply the purge gas to Nafion dryers.Common methods include those with no consumables, likereusing the sample gas itself after it is partially dried passingthrough the inner Nafion membrane (as in this study) or sup-plying purge air from a dry-air generator, and methods withconsumables that must be replaced, such as using molecularsieve to remove all the water from the sample after the Nafionand before it is reused as the purge gas (e.g. Stephens et al.,2011) or dry air from a tank. The choice depends largely onwhat the tolerance is for residual water in the sample gas andhow frequently technicians are able to service the dryer.

For the Earth Networks installations, we designed a sim-ple drying inlet system for ambient air monitoring (Fig. 1)using a 72-inch-long Nafion membrane dryer (PermaPure,Inc., model MD-050-72S-1). This inlet drying configurationtakes advantage of the extra capacity of the external analyzerpump to redirect 30 % of the dried gas exiting the Nafion tothe outer shell side of the dryer, creating both a gradient in theH2O partial pressure and total gas pressure across the Nafionmembrane. This total pressure drop across the membrane en-hances the drying capacity of the Nafion and conserves sam-ple and reference gas volumes.

This setup also unavoidably produces partial pressure gra-dients in CO2 and CH4 across the membrane that may allowsmall amounts of these trace gases and the major componentsof air to also permeate across the membrane (Ma and Skou,2007). To reduce the influence of any such permeation on thesample measurements, the network sampling setup uses ac-tive pressure stabilization for all ambient air intakes and ref-erence tank gases. This will ensure that the Nafion is exposedto sample and reference gases at identical total pressures, sothat any direct effect of pressure-dependent permeation ofCO2 and CH4 is canceled in the comparison between sampleand reference gases.

Even with the pressure stabilization, one important differ-ence still exists between sample air and reference tanks. Thesample gas enters the Nafion with much higher moisture lev-els than the dry reference gases. The permeation of CO2 and

Atmos. Meas. Tech., 6, 1217–1226, 2013 www.atmos-meas-tech.net/6/1217/2013/

L. R. Welp et al.: Design and performance of a Nafion dryer 1219

P

F

F

F

F

F

F

F

F

F

FF

F

F

FLASKPORT

AIRLINE#1

STDGAS#1

AIRLINE#3

AIRLINE#2

STDGAS#4

STDGAS#3

STDGAS#2

SPARE

PUMPMODULE

#1

VENT~ 10 L/mLINEAR

OSCILLATING“AQUARIUM”

PUMP(ALMA ALGSA)

MKS 640PRESSURE

CONTROLLER

PRESSURESENSOR

VALCO 8-PORTSELECTOR

VALVE

~ 600 TORRNAFION 72”

MD-050-725-1

~ 30 scc/m< 10 TORR

VACUUM PUMP(VACUUBRAND

MZ2NT)

~ 70 scc/m

< 10 TORR

CRDS OPTICAL CELL

PICARRO 2301

SAMPLE MODULE

VENT

FILTER

CRITICALORIFICE

LEGEND

~~

Fig. 1.The gas-handling configuration for the Earth Networks greenhouse gas monitoring stations. The “Sample Module” houses the Nafiondryer system. All sample air and reference gases pass through the Nafion dryer.

CH4 through the membrane may be moisture-dependent (Maet al., 2005), and this could lead to differential biases betweensample and calibration gases that do not cancel.

The ideal inlet drying system removes water vapor with-out modifying CO2 or CH4. In this study, we test the per-formance of the inlet drying system by ensuring that anychanges to CO2 or CH4 are small or correctable. We presentresults of tests that evaluate the performance of two simi-lar drying systems, a lab-built test setup and an Earth Net-works sample module. We show that the dryer is effective inreducing H2O to between 0.1 to 0.15 % levels, and quanti-fied moisture-dependent changes in the permeation of CO2and CH4 across the Nafion membrane. We evaluated the labtest setup before and after 9 months of continuous opera-tion at room temperature to test for aging effects, and testedthe Earth Networks sample module at room temperature andheated. We also quantify the time required for the Earth Net-works Nafion dryer sample module to stabilize after dry ref-erence gases are introduced and test for any transient offsetsthat may occur.

2 Methods

Testing the Nafion dryer requires a humidified air sourcewith known dry-gas mixing ratios of CO2 and CH4. It isdifficult to humidify air without changing its dry-gas CO2mixing ratio slightly because the gasses dissolve in water

via Henry’s Law and also the propensity of CO2 to inter-act with water adsorbed on tubing walls. For this reason,we compared our Nafion dryer system with a cryotrap at−97◦C using the experimental design in Fig. 2. Similar cry-otraps have been used in this community for decades andany effects on CO2 and CH4 are known to be very small,but published studies documenting the cryotrap biases arelacking. We confirmed this in our laboratory comparing drytank air to the same air that was humidified to 2 % H2O by abubbler filled with 15 mL of water acidified with phospho-ric acid at 20◦C and then dried by a cryotrap at−97◦C.Initially, when the cryotrap is placed in the chiller, it ab-sorbs CO2 onto surfaces, but quickly saturates and achievesa steady state. The time that it takes to reach steady statedepends on the surface area inside the trap (walls and glassbeads). Henry’s Law predicts an increase in the CO2 of thehumidified/dried treatment of∼ 0.005 ppm based on the re-lease of dissolved CO2 from the water that evaporates. Werepeated the experiment twice and found differences closeto Henry’s Law, 0.004± 0.002 ppm and 0.006± 0.009 ppm.Likewise, the Henry’s Law prediction for CH4 is∼ 0.001 ppband we found differences within measurement uncertainty of0.003± 0.014 ppb and−0.038± 0.06 ppb. Thus, we believeshowing that the Nafion system is as good as such a cryotrapis a suitable performance benchmark.

The analyzer used for testing this application was a Pi-carro G2301 CRDS CO2/CH4/H2O gas analyzer, but this

www.atmos-meas-tech.net/6/1217/2013/ Atmos. Meas. Tech., 6, 1217–1226, 2013

1220 L. R. Welp et al.: Design and performance of a Nafion dryer

F F

F

VACUUMPUMP

-97° CCRYOTRAP

BUBBLER

TANK

VACUUMPUMP

FLOWMETER

PICARRO 2301MKS640

PRESSURECONTROLLER

~ 600TORR

CRDS OPTICAL CELL

NAFION 72”

MD-050-725-1

~ 30 scc/m< 10 TORR

VENT

~ 70 scc/m

< 10 TORR

CRDSCHANGE-OVERVALVE

BUBBLERCHANGE-OVERVALVE

-97° CCRYOTRAP

~ 70 scc/m

~ 100 scc/m ~ 70 scc/m

~ 30 scc/m

F

CRITICALORIFICE

FILTER

LEGEND

Fig. 2. Experimental test setup for testing the Nafion dryer. Dry tank air is humidified with a bubbler. The Nafion and cryotrap treatmentsare continuously flowing and the CRDS changeover valve allows the CRDS analyzer to alternately monitor each one. The second cryotrapdownstream of the CRDS changeover valve was added after Exp. 1.1 and 1.2 to remove the residual water exiting the Nafion and eliminatethe need to apply any water vapor correction. In the dry-air experiments the first cryotrap was either left at room temperature or removed. Thedownstream cryotrap was used to remove the residual water from the Nafion treatment, thus avoiding any H2O/CO2 interaction on surfaceswithin and immediately upstream of the Picarro.

application may be used with any similar gas analyzer. Theexternal pump used was a KNF pump unit supplied by Pi-carro, retrofit with tighter leak-free connections in order toensure that the gas flow exiting the pump was identical to thegas flow through the CRDS analyzer. The same analyzers areused in the Earth Networks greenhouse gas network designbut with a different external pump and a few slight changesto set air flow rates. Flow rates in the test system were con-trolled by manually adjusting needle valves at the CRDS inletand in the Nafion counterflow loop (Fig. 2). In the deployednetwork design, the flow rates are set by replacing the Picarrofactory default O’Keefe A-18-NY orifice with a smaller A-9-NY, and in the Nafion counterflow by an A-6-NY orifice(Fig. 1). For these orifices to work, the downstream pressurehas to be low enough to assure sonic velocity in the orifice.This is needed for the critical flow to be constant regardlessof upstream pressure. It is also important to filter the gas up-stream of the orifice to avoid clogging.

Dry air from a compressed gas cylinder was passedthrough a temperature-controlled bubbler-style humidifierin wet-air tests or bypassed in dry-air tests (Fig. 2).The resulting wet air (or dry air) was split between theNafion dryer and the cryotrap, each constantly flowing at∼ 70 cm3 min−1, and alternately sampled by the Picarro an-alyzer, or sent to a waste pump. The primary cryotrap was

stainless steel, with a 1/2 inch outer tube and a 1/8 inch innerdip tube. Air flowed in through the outer tube, and 3 mm glassbeads at the bottom reduced the chance of ice crystals enter-ing the inner tube along with the airflow out of the cryotrap.We also added heat tape to the top of the cryotrap, above thechiller, to discourage liquid water build up at the top of thetrap. The H2O mixing ratio of the humidified air could beadjusted by changing the temperature of the water bath sur-rounding the humidifier and by changing the gas pressure in-side the humidifier. Delivery pressure to the Nafion dryer andcryotrap was maintained at 600 Torr by an MKS640 pressurecontroller in order to simulate conditions for routine sam-pling of ambient air.

The experiments using this test setup of the Nafion dry-ing system were conducted at ambient laboratory tempera-ture of∼ 25± 1◦C over the course of each experiment. Thedrying systems deployed by Earth Networks are in an en-closure warmed approximately 10–20◦C above ambient inorder to minimize the chance of condensation forming in theinlet system. This temperature increase makes the Nafion lessefficient at removing water (Leckrone and Hayes, 1997), byapproximately 0.01 % H2O, based on side-by-side operationof the test system described here and a heated network sys-tem sampling from the same air intake from the Scripps In-stitution of Oceanography (SIO) pier in La Jolla, California.

Atmos. Meas. Tech., 6, 1217–1226, 2013 www.atmos-meas-tech.net/6/1217/2013/

L. R. Welp et al.: Design and performance of a Nafion dryer 1221

Table 1.Mid-point differences between Nafion and cryotrap treatments (Nafion minus the mean of cryotrap measurements before and after).

CO2 CH4

H2O Temp na mean stdevb mean stdevb

Date Exp Setup (%) (◦C) (ppm) (ppm) (ppb) (ppb)

29-Jul-11 1.1c Test setup 0 25 11 −0.01 0.01 −0.07 0.271-Aug-11 1.2c Test setup 2.1 25 4 −0.03 0.01 −0.08 0.052-Aug-11 1.3 Test setup 2.1 25 5−0.05 0.01 −0.02 0.1018-Aug-11 1.4 Test setup 2.1 25 3−0.03 0.01 0.10 0.0524-May-12 2.1c Test setup 0 25 14 0.01 0.01 −0.05 0.1030-May-12 2.2 Test setup 2.2 25 10−0.03 0.01 0.08 0.0623-Feb-13 3.1 Sample module 0 25 22−0.02 0.01 0.03 0.1118-Feb-13 3.2 Sample module 2.0 25 7−0.07 0.01 −0.01 0.1519-Feb-13 3.3d Sample module 2.0 25 4 −0.06 0.01 0.11 0.0822-Feb-13 3.4 Sample module 0 35 6−0.04 0.01 0.08 0.1520-Feb-13 3.5 Sample module 2.0 35 7−0.10 0.01 0.22 0.13

a number of Nafion intervals measured;b standard deviation of the mid-point differences;c water vapor correction applied to the Nafiontreated data;d ethanol and dry ice slush (−72◦C) used in place of the−97◦C chiller.

The permeability of the Nafion to CO2 has also been shownto be temperature-dependent (Ma and Skou, 2007). For thesereasons, we repeated the testing using one of the Earth Net-works sample modules described in Fig. 1 at room tempera-ture and also heated.

Three types of experiments were completed at the startof the test setup operation: one dry-air run (Exp. 1.1), onewet-air run (2.1 % H2O, Exp. 1.2), and two wet-air runs(2.1 % H2O) with a secondary cryotrap after the Nafion dryerto eliminate the need to apply the water vapor correction(Exp. 1.3 and 1.4). The secondary cryotrap was placed im-mediately before the CRDS. This secondary cryotrap used a“cold finger” design similar to that of the primary cryotrap,but smaller, 1/4 inch stainless steel outer tube and 1/16 inchinner tube, with no glass beads.

In all these experiments, gas was constantly flowingthrough the Nafion system at∼ 100 cm3 STP min−1 (with∼ 30 cm3 STP min−1 of that redirected to the counterflowpurge) and the cryotrap at∼ 70 cm3 STP min−1. We used theCRDS changeover valve (Fig. 2) to alternately switch theCRDS intake between the Nafion and cryotrap, quantifyingany differences in the CO2 and CH4 mixing ratios betweenthe two treatments.

For the wet-air runs, the H2O values were approximately2 %, as high as we could achieve with the bubbler humidi-fier at 20◦C, maintaining an operating pressure greater than600 Torr, and keeping the test setup at laboratory tempera-ture. This corresponds to a dew point temperature of approx-imately 17◦C. Since the Nafion-dryer treatment without thesecondary cryotrap did not completely remove all of the H2Ofrom the air, and in the case of the dry-air run it added H2Oto the air, we applied the water vapor correction from Chen etal. (2010) as recommended by Rella et al. (2013) to the data

from Exps. 1.1 and 1.2. (Appendix A). These corrections areon the order of 1.0 ppm CO2 and 3.7 ppb CH4 for 0.2 % H2O.

For most of these tests, we switched the changeover valveevery 60 min. Preliminary tests using 15 min switching in-tervals showed drifting CO2 values presumably because ofdifferent flow resistances between the Nafion system and thecryotrap, causing small pressure changes that in turn causedCO2 to be adsorbed or desorbed on the walls after switching.

After completing a first round of tests as described above,the Nafion dryer test setup inlet system with the CRDS wasused continuously for nine months of routine air measure-ments. We then repeated the dry-gas experiment (Exp. 2.1)and wet-gas experiment (Exp. 2.2) to see if the aging of theNafion over the nine-month period had any impact on thepermeability of CO2 and CH4.

For testing the Earth Networks sample module, we madeone modification, splitting the flow of air downstream of thepressure controller to send a portion through a bypass or cry-otrap treatment. The secondary cryotrap was used in all ofthese experiments (Exp. 3.1–3.5).

All data were processed on 1 min running means, discard-ing the first 30 min of data after switching. There was noevidence of drift in CO2 or CH4 in the last 30 min of datafor each treatment. The number of switching intervals var-ied in each experiment, as the primary cryotrap would plugwith ice after several hours of use, thereby blocking the airflow and ending the experiment. Biases between the Nafionand cryotrap treatments were calculated as the differencesbetween Nafion and the mean of the two cryotrap treatmentsbefore and after, i.e. the mid-point difference, and then av-eraged over the total number of sampling pairs. In Table 1,we report the mean and standard deviation of the mid-pointdifferences.

www.atmos-meas-tech.net/6/1217/2013/ Atmos. Meas. Tech., 6, 1217–1226, 2013

1222 L. R. Welp et al.: Design and performance of a Nafion dryerC

O2 (

ppm

)

393.2

393.3

393.4

CH

4 (pp

b)

elapsed hours0 1 2 3 4 5 6 7 8 9 10

1873.0

1873.5

1874.0

a

b

393.5

Fig. 3. Wet-air Exp. 1.3 on 2 August 2011. The cryotrap treatmentis shown in black and the Nafion treatment is in grey for(a) CO2,(b) CH4. The first 30 min of data were excluded for each treatment.The straight lines are mean values for the cryotrap (black dashed)and Nafion (solid grey) over the entire experiment. The cryotrapsplugged up with ice, ending the experiment after approximately11 h. The secondary cryotrap was used in this experiment to removethe water remaining after the Nafion, so no water vapor correctionwas applied.

3 Results

3.1 Test setup

In the first dry-air experiment, Exp. 1.1 (Fig. S1 in the Sup-plement), we ran dry tank air with a CO2 dry-gas mixing ratioof ∼ 506 ppm and CH4 of ∼ 4787 ppb, switching between theNafion and bypass. In this case, the cryotrap was in-line butnot chilled, rather at ambient laboratory temperature sincethe air was dry already. The mean mid-point difference ofthe Nafion and bypass treatments was−0.01± 0.01 ppm forCO2 and−0.07± 0.27 ppb for CH4 (summarized in Table 1).Negative values mean that there was slightly less CO2 andCH4 after the Nafion treatment compared with the cryotraptreatment.

In the first wet-air experiment, Exp. 1.2 (Fig. S2), wehumidified tank air with a CO2 dry-gas mixing ratio of∼ 393 ppm CO2 and ∼ 1874 ppb CH4 to 2.1 % H2O. TheNafion dryer reduced the H2O in the humidified sample downto 0.12 %, and we applied the water vapor correction to thesedata to calculate dry-gas mixing ratios for CO2 and CH4using Eqs. (A1) and (A2). We then compared the Nafiondryer treatment with the cryotrap and found a mean mid-point difference over four pairs of Nafion/cryotrap switchingof −0.03± 0.01 ppm for CO2 and−0.08± 0.05 ppb for CH4(Table 1).

The final two wet-air experiments, Exp. 1.3 (Fig. 3) andExp. 1.4 (Fig. S3), used the additional cryotrap to com-pletely remove the residual water that is not removed by theNafion. The CRDS confirmed that the air was dry and no wa-ter vapor corrections were applied. The largest mean Nafionminus cryotrap differences were−0.05± 0.01 ppm of CO2and 0.1± 0.05 ppb of CH4 (Table 1). In this set of wet-airexperiments, the cold trap plugged with ice after 3 to 5 pairsof Nafion/cryotrap switching.

3.2 Test setup after nine months of operation

After nine months of continuously monitoring ambient airfrom intakes on the SIO pier, we repeated similar tests tosee if age or use affected the permeability of the Nafion toCO2 and CH4. We repeated the dry-air experiment, Exp. 2.1,using a different dry-air tank with approximately∼ 394 ppmCO2 and ∼ 1873 ppb of CH4, again, leaving the cryotrapsat room temperature (Fig. S4). We applied the water vaporcorrection to account for the small amount of water addedfrom the Nafion treatment. The mean of 14 pairs of Nafionminus cryotrap mid-point differences was 0.01± 0.01 ppmfor CO2 and−0.05± 0.1 ppb for CH4 (Table 1).

We also repeated the wet-air experiment, Exp. 2.2(Fig. S5) using the same dry air tank as in Exp. 1.2, buthumidified to 2.2 % H2O. We used the secondary cryotrap,which eliminated the need to apply the water vapor cor-rection. The mean of 10 pairs of Nafion minus cryotrapmid-point differences was−0.03± 0.01 ppm of CO2 and0.08± 0.06 ppb of CH4 (Table 1). These results are similarto Exp. 1.3 and 1.4, run prior to routine operation, and showno sign of age effects on Nafion after nine months.

3.3 Earth Networks sample module

In early 2013, we tested a typical Earth Networks samplemodule, inlet dryer system, to compare its performance tothe test setup system and further evaluate the Nafion at el-evated temperature. Several sets of experiments were donewith the sample module, and it became clear in these exper-iments that CO2 mixing ratios were influenced by exchangeprocesses on surfaces inside the stainless steel tubing andtraps with glass beads. Replacing tubing and cleaning thetraps and beads led to differences in the Nafion minus cry-otrap offsets. These differences highlighted the challenges ofmeasuring CO2 in wet air as CO2 and H2O adsorb to sur-faces and compete for the same surface sites. Representa-tive results are summarized in Table 1. In the dry-air roomtemperature experiments (∼ 25◦C, Exp. 3.1, Fig. S6), theNafion minus cryotrap offset was−0.02± 0.01 ppm for CO2and 0.03± 0.11 ppb for CH4. Two experiments were doneusing wet air (2.0 % H2O, Exp. 3.2, Fig. S7 and Exp. 3.3,Fig. S8) at room temperature, the largest offsets increased by0.05 ppm to−0.07± 0.01 ppm for CO2 and only negligibly(0.11± 0.08 ppb) for CH4. The sample module dried the airfrom 2.0 % to 0.05 % at room temperature.

The drying capacity and the CO2 permeability wereslightly different in the sample module experiments com-pared with the test setup. This is because the flow rates andpressures inside the Nafion and in the purge counterflow areslightly different using the critical orifices to regulate theflow compared to the needle valves in the test setup. This setsup different gradients in partial pressures across the Nafionmembrane resulting in different rates of drying and CO2 per-meation.

Atmos. Meas. Tech., 6, 1217–1226, 2013 www.atmos-meas-tech.net/6/1217/2013/

L. R. Welp et al.: Design and performance of a Nafion dryer 1223

393.15

393.20

393.25

393.30

393.35

1874.5

1875.0

1875.5

1876.0

1876.5

1877.0

minutes0 10 20 30 40 50 60

minutes

CO

2 (pp

m)

0 10 20 30 40 50 60

a b

CH

4 (pp

b)Fig. 4. Stabilization times for CO2 and CH4 of dry reference tanksafter switching from moist ambient air. These data were collectedduring the month of January 2013 using the Earth Networks stationat SIO with a sample module Nafion dryer inlet. Grey lines are the1 min averages of the first reference tank switched to after ambientair during the daily reference tank checks for(a) CO2 and(b) CH4.This tank was sampled for 1 h. The 5 min ensemble averages overevery daily tank run are shown in black, plotted at the mid-point,along with 1-standard deviation error bars.

The high temperature experiments, approximately 35◦C,dry and wet (Exp. 3.4, Fig. S9 and Exp. 3.5, Fig. S10),showed larger losses of CO2 across the Nafion membraneas predicted. The dry air Nafion minus cryotrap differenceswere −0.04± 0.01 ppm for CO2 and 0.08± 0.15 ppb forCH4. The wet air (2.0 %) differences were−0.10± 0.01 ppmfor CO2 and 0.22± 0.13 ppb for CH4. The drying efficiencyof the Nafion was reduced, yielding 0.10 % H2O at the highertemperature.

3.4 Transition times

The experiments described above (Sects. 3.1, 3.2 and 3.3)quantified the offsets in CO2 and CH4 between Nafion andcryotrap treatments in steady-state wet and dry conditions.We were also interested in addressing how much time is re-quired for the system to stabilize under routine conditionswhen the intake selector switches from wet ambient air to dryreference tank air. It is possible that switching from moist airto a dry reference tank could cause transient offsets in CO2and CH4 while the Nafion dries. In Fig. 4 we show the ambi-ent air to dry tank transitions from one month of routine oper-ation by the SIO Earth Networks station in January 2013. Thewater vapor correction in Appendix A was applied to thesedata. The humidity of the reference gas exiting the Nafion isshown to decrease from about 0.13 % to 0.09 % as the Nafionslowly dries over the course of one hour. After switching tothe dry tank air, the Nafion inlet sample module is initiallya weak sponge for CO2, and it takes the CRDS measure-ments about 10 min to recover and stabilize. The CO2 5 minmean standard deviations decrease from 0.37 to 0.20 ppmfrom 10 min to 60 min after tank switching with a 0.005 ppmincrease in the mean values. The CH4 5 min mean standarddeviation remains nearly constant at 0.0003 ppb CH4 overthis hour with negligible drift in the mean values.

4 Discussion

We find that the permeation of CO2 through the Nafionmembrane in the wet-air experiments, relative to a−97◦Ccryotrap, results in offsets of−0.07 ppm at 25◦C and−0.10 ppm and 35◦C using the Earth Networks sample mod-ule configuration. As expected from previous literature (Maet al., 2005; Ma and Skou, 2007), we found that the CO2loss across the Nafion increased with sample air humidityand with temperature. The permeation rates that we foundare far less than the maximum permeation rates measured byMa and Skou (2007). The maximum permeability they foundin the fully hydrated Nafion case (soaked in liquid water)at 50◦C translates to a 2.2 ppm drop in CO2 through thedryer as a result of loss across the membrane. Our results aremuch closer to their dry Nafion condition, which we estimatewould be a 0.02 ppm loss at 25◦C and a 0.031 ppm loss at45◦C. Their experiments used argon as the gas matrix insteadof air. In the tested Earth Networks measurement design, thesample modules were not temperature controlled, but ratherheated above ambient to prevent condensation. Partly as aresult of this work, Earth Networks is modifying the sam-ple module design to be thermostated in the 35◦C to 40◦Crange. This will reduce vulnerability to different CO2 lossrates through the Nafion at different installations and timesof year.

The Nafion effects on CH4 are much smaller, and mixed,than CO2 because CH4 is not a polar molecule. The offsetsranged from a loss of 0.07 ppb to a gain of 0.20 ppb. Thesevalues are close to the detection limit of the CRDS analyzer.

Our procedure of delivering both sample air and referencegases through the Nafion at the same pressure makes the dry-ing bias due to permeation through the Nafion smaller in rou-tine operation, i.e. the bias will cancel out after applying theCO2 and CH4 calibrations to the sample air based on ref-erence gas analysis. For example, in our tests of the heatedsample module, we saw a loss of 0.10 ppm CO2 through theNafion with wet air and a loss of 0.04 ppm with dry air. Afterapplying the CO2 calibration to the sample air, this would re-sult in a 0.05 ppm systematic loss of CO2 (avoiding roundingerrors). This is the value we use in our uncertainty analy-sis. Also, because the Nafion dryer has a slow response timefor H2O, the Nafion humidifies the reference gases so theuncertainty in the water vapor correction further cancels out(Richardson et al., 2012). This effect could be capitalized onfurther by running each reference gas separately for a shorterperiod of time (e.g. 20–30 min, rather than multiple tanks runsequentially) to limit the degree to which the Nafion driesout.

The advantage of partially drying the sample gas canbe determined by comparing the estimated measurementuncertainties from multiple sources for three cases: (1) us-ing the manufacturer-supplied water vapor correction co-efficients, (2) using instrument-specific correction coeffi-cients experimentally determined at the start of operation

www.atmos-meas-tech.net/6/1217/2013/ Atmos. Meas. Tech., 6, 1217–1226, 2013

1224 L. R. Welp et al.: Design and performance of a Nafion dryer

Table 2.Summary of errors for different CRDS application approaches.

No dryingNo drying with with instrument- Nafion drying withuniversal water specific water universal water

vapor correctiona vapor correction vapor correctiona

2 % H2O 2 % H2O 0.15 % H2O

CO2 CH4 CO2 CH4 CO2 CH4Error types (ppm) (ppb) (ppm) (ppb) (ppm) (ppb)

Instrument precisionb 0.025 0.22 0.025 0.22 0.025 0.22Noise in H2Oc 0.015 0.071 0.015 0.071 0.014 0.069Water vapor correctiond 0.10 1.6 0.06 0.7 0.015 0.21Nafion-induced biase N/A N/A N/A N/A 0.05 0.22Sum of errorsf 0.10 1.62 0.07 0.74 0.08 0.53

a Universal water vapor correction used was from Chen et al. (2010).b 5 min mean from the Picarro G2301 manufacturerspecifications sheet downloaded 18 October 2012.c Based on Rella et al. (2013).d Based on Rella et al. (2013).e From thisexperiment.f The quadrature sum of errors for first 3 rows plus the Nafion bias. Error= sqrt(a2

+ b2+ c2) + d. This was

used assuming that the first 3 errors are not correlated and do not act in the same direction.

and (3) partial drying with our Nafion dryer system and ap-plying manufacturer-supplied water vapor corrections to theresidual water (Table 2). Instrument precision from the man-ufacturer’s specifications for 5 min averages is 0.025 ppm forCO2 and 0.22 ppb for CH4, with precision improving withlonger integration times. The random noise in the instru-ment’s H2O measurement contributes uncertainty to CO2 andCH4 through the water vapor correction as described in Rellaet al. (2013). At 2 % H2O, it adds uncertainties of 0.015 ppmfor CO2 and 0.071 ppb for CH4. At 0.15 %, it is reducedonly slightly to 0.014 ppm for CO2 and 0.069 ppb for CH4.Rella et al. (2013) summarized the uncertainty in the watervapor correction using the results from several different in-struments tested in different laboratories. The uncertainty at2 % H2O is approximately 0.10 ppm for CO2 and 1.6 ppb forCH4 using the coefficients from Chen et al. (2010) and is re-duced to 0.06 ppm for CO2 and 0.7 ppb for CH4 using experi-mentally determined instrument-specific coefficients. Dryingthe sample air to 0.15 % H2O lowers the water vapor correc-tion uncertainty using the Chen et al. (2010) coefficients to0.015 ppm for CO2 and 0.21 ppb for CH4.

The total estimated uncertainty in the CO2 and CH4 mea-surements at 2 % H2O is determined by adding the instru-ment precision, the effect of noise in the H2O measurements,and the uncertainty in the water vapor correction (first threerows in Table 2) in quadrature because they do not neces-sarily act in the same direction, and then add the Nafion-induced bias on top of that. The estimated errors for uti-lizing the Nafion dryer and correcting the residual waterusing the manufacturer-supplied coefficients based on Chenet al. (2010) are 0.08 ppm for CO2 and 0.41 ppb for CH4.This is an improvement over not drying the air and using theChen et al. (2010) coefficients, and comparable to character-izing an instrument-specific water vapor correction (Table 2).Another study by Nara et al. (2012) also recommends at least

partial drying of the sample air for precise measurements.The statistical advantages of applying smaller water vaporcorrections to partially dried air increase as water levels ex-ceed 2 %, however, these advantages may be offset by greaterpermeation of CO2 through the Nafion at higher humidity.

The errors discussed in Rella et al. (2013) are limited intwo ways. First, they have not attempted to diagnose theresidual systematic error from CO2 and H2O adsorption ontubing walls. Especially in the case of the water droplet meth-ods where H2O is continually changing, this can lead to bi-ases in the CO2. It is hard to prove the absence of a bias atthe level of 0.1 ppm at 2 % H2O. The random errors acrossthe different water vapor correction experiments in Rella etal. (2013) were found to be less than 0.1 ppm up to severalpercent H2O, but this does not confirm the absence of sys-tematic errors. Second, at least part of the range in the errorsof manufacturer-supplied water vapor correction coefficientsreported from the participating laboratories could be the re-sult of experimental artifacts and may not reflect actual re-sponse differences among analyzers.

The main systematic error in our case is the cryotrap bias,which the community has accepted as small enough for manyyears and we confirmed is negligible. It is also possible thatindividual Nafion dryers have different permeability charac-teristics, and more Nafion dryers (e.g. beyond the two testedhere) may need to be tested.

5 Conclusions

We have tested a design for drying air using a Nafion dryer,both as a test setup and as a field sample module, in conjunc-tion with a CRDS analyzer. These inlet dryer systems areentirely self-sustaining and have required no intervention inover a year of operation. We also find that the performance of

Atmos. Meas. Tech., 6, 1217–1226, 2013 www.atmos-meas-tech.net/6/1217/2013/

L. R. Welp et al.: Design and performance of a Nafion dryer 1225

the system is not subject to problems of Nafion degradation,at least over a 9-month time frame. Our experiments showthat the bias between the permeability of CO2 through theNafion membrane in wet and dry tank air may be as largeas 0.05 ppm CO2 using an Earth Networks sample moduleconfiguration. The bias is negligible for CH4. Combining theNafion effects with other random errors in the CRDS mea-surements, we estimate the uncertainty to be approximately0.08 ppm for CO2 and 0.53 ppb for CH4.

The setup eliminates the need to establish the water va-por correction on each analyzer and monitor their stabilityover time. It also reduces the complexity of post-processingthe data if the correction is found to change over time. Thismethod more fully capitalizes on the ability of the CRDS toprovide high-quality measurements with reduced calibrationactivities, thereby saving labor costs for deployments involv-ing an extensive network of analyzers such as that plannedby Earth Networks.

Appendix A

Water vapor correction

The water vapor correction from Chen et al. (2010), as rec-ommended by Rella et al. (2013), is summarized in Eqs. (A1)and (A2).

(CO2)wet

(CO2)dry= 1+ aHrep+ bH 2

rep (A1)

(CH4)wet

(CH4)dry= 1+ cHrep+ dH 2

rep (A2)

whereHrep is the reported H2O mixing ratio by the analyzer,(CO2)wet and (CH4)wet are the measured mixing ratios ofthe wet gas, (CO2)dry and (CH4)dry are the true dry-gas mix-ing ratios,a = −0.012000,b = −0.0002674,c = −0.00982,andd = −0.000239.

Supplementary material related to this article isavailable online at:http://www.atmos-meas-tech.net/6/1217/2013/amt-6-1217-2013-supplement.pdf.

Acknowledgements.The authors thank Chris Rella and Aaron VanPelt at Picarro, Inc. for their discussions, technical assistance, andconstructive comments on this manuscript. Earth Networks, Inc.funded this work.

Edited by: O. Tarasova

References

Baer, D. S., Paul, J. B., Gupta, M., and O’Keefe, A.: Sensitive ab-sorption measurements in the near-infrared region using off-axisintegrated-cavity-output spectroscopy, Appl. Phys. B, 75, 261–265,doi:10.1007/s00340-002-0971-z, 2002.

Bakwin, P., Tans, P., Hurst, D., and Zhao, C.: Measurements of car-bon dioxide on very tall towers: results of the NOAA/CMDL pro-gram, Tellus B, 50, 401–415, 1998.

Butler, M. P., Davis, K. J., Denning, A. S., and Kawa, S. R.: Us-ing continental observations in global atmospheric inversions ofCO2: North American carbon sources and sinks, Tellus B, 62,550–572, 10.1111/j.1600-0889.2010.00501.x, 2010.

Chen, H., Winderlich, J., Gerbig, C., Hoefer, A., Rella, C. W.,Crosson, E. R., Van Pelt, A. D., Steinbach, J., Kolle, O., Beck,V., Daube, B. C., Gottlieb, E. W., Chow, V. Y., Santoni, G. W.,and Wofsy, S. C.: High-accuracy continuous airborne measure-ments of greenhouse gases (CO2 and CH4) using the cavity ring-down spectroscopy (CRDS) technique, Atmos. Meas. Tech., 3,375–386,doi:10.5194/amt-3-375-2010, 2010.

Crosson, E. R.: A cavity ring-down analyzer for measuring atmo-spheric levels of methane, carbon dioxide, and water vapor, Appl.Phys. B, 92, 403–408,doi:10.1007/s00340-008-3135-y, 2008.

Gurney, K., Law, R., Denning, A., Rayner, P., Baker, D., Bousquet,P., Bruhwiler, L., Chen, Y., Ciais, P., Fan, S., Fung, I., Gloor,M., Heimann, M., Higuchi, K., John, J., Maki, T., Maksyutov,S., Masarie, K., Peylin, P., Prather, M., Pak, B., Randerson, J.,Sarmiento, J., Taguchi, S., Takahashi, T., and Yuen, C.: Towardsrobust regional estimates of CO2 sources and sinks using atmo-spheric transport models, Nature, 415, 626–630, 2002.

Houweling, S., Aben, I., Breon, F.-M., Chevallier, F., Deutscher,N., Engelen, R., Gerbig, C., Griffith, D., Hungershoefer, K.,Macatangay, R., Marshall, J., Notholt, J., Peters, W., and Serrar,S.: The importance of transport model uncertainties for the esti-mation of CO2 sources and sinks using satellite measurements,Atmos. Chem. Phys., 10, 9981–9992,doi:10.5194/acp-10-9981-2010, 2010.

Leckrone, K. J. and Hayes, J. M.: Efficiency and temperature de-pendence of water removal by membrane dryers, Analyt. Chem.,69, 911–918,doi:10.1021/ac9610220, 1997.

Lin, J. C., Gerbig, C., Wofsy, S. C., Daube, B. C., Matross, D. M.,Chow, V. Y., Gottlieb, E., Andrews, A. E., Pathmathevan, M., andMunger, J. W.: What have we learned from intensive atmosphericsampling field programmes of CO2?, Tellus Ser. B, 58, 331–343,doi:10.1111/j.1600-0889.2006.00202.x, 2006.

Ma, S. and Skou, E.: CO2 permeability in Nafion®EW1100at elevated temperature, Solid State Ionics, 178, 615–619,doi:10.1016/j.ssi.2007.01.030, 2007.

Ma, S., Odgaard, M., and Skou, E.: Carbon dioxide permeabilityof proton exchange membranes for fuel cells, Solid State Ionics,176, 2923–2927,doi:10.1016/j.ssi.2005.09.024, 2005.

Nara, H., Tanimoto, H., Tohjima, Y., Mukai, H., Nojiri, Y.,Katsumata, K., and Rella, C. W.: Effect of air composition(N2, O2, Ar, and H2O) on CO2 and CH4 measurement bywavelength-scanned cavity ring-down spectroscopy: calibrationand measurement strategy, Atmos. Meas. Tech., 5, 2689–2701,doi:10.5194/amt-5-2689-2012, 2012.

NRC: Verifying greenhouse gas emissions: Methods to support in-ternational climate agreements, 124, 2010.

www.atmos-meas-tech.net/6/1217/2013/ Atmos. Meas. Tech., 6, 1217–1226, 2013

1226 L. R. Welp et al.: Design and performance of a Nafion dryer

Rella, C. W.: Accurate greenhouse gas measurements in humid gasstreams using the Picarro G1301 carbon dioxide/methan/watervapor gas analyzer, 1–18, 2010.

Rella, C. W., Chen, H., Andrews, A. E., Filges, A., Gerbig, C.,Hatakka, J., Karion, A., Miles, N. L., Richardson, S. J., Stein-bacher, M., Sweeney, C., Wastine, B., and Zellweger, C.: Highaccuracy measurements of dry mole fractions of carbon diox-ide and methane in humid air, Atmos. Meas. Tech., 6, 837–860,doi:10.5194/amt-6-837-2013, 2013.

Richardson, S. J., Miles, N. L., Davis, K. J., Crosson, E. R., Rella, C.W., and Andrews, A. E.: Field testing of cavity ring-down spec-troscopy analyzers measuring carbon dioxide and water vapor, J.Atmos. Ocean. Technol., 29, 397–406,doi:10.1175/JTECH-D-11-00063.1, 2012.

Stephens, B. B., Gurney, K. R., Tans, P. P., Sweeney, C., Pe-ters, W., Bruhwiler, L., Ciais, P., Ramonet, M., Bousquet, P.,Nakazawa, T., Aoki, S., Machida, T., Inoue, G., Vinnichenko,N., Lloyd, J., Jordan, A., Heimann, M., Shibistova, O., Lan-genfelds, R. L., Steele, L. P., Francey, R. J., and Denning, A.S.: Weak northern and strong tropical land carbon uptake fromvertical profiles of atmospheric CO2, Science, 316, 1732–1735,doi:10.1126/science.1137004, 2007.

Stephens, B. B., Miles, N. L., Richardson, S. J., Watt, A. S.,and Davis, K. J.: Atmospheric CO2 monitoring with single-cell NDIR-based analyzers, Atmos. Meas. Tech., 4, 2737–2748,doi:10.5194/amt-4-2737-2011, 2011.

Winderlich, J., Chen, H., Gerbig, C., Seifert, T., Kolle, O., Lavric,J. V., Kaiser, C., Hofer, A., and Heimann, M.: Continuous low-maintenance CO2/CH4/H2O measurements at the Zotino TallTower Observatory (ZOTTO) in Central Siberia, Atmos. Meas.Tech., 3, 1113–1128,doi:10.5194/amt-3-1113-2010, 2010.

WMO: Report of the 16th WMO/IAEA meeting on carbon dioxide,other greenhouse gases, and related measurement techniques,Wellington, New Zealand, 25–28 October 2011, GAW ReportNo. 194, WMO TD No. 1553, 67, 2012.

Atmos. Meas. Tech., 6, 1217–1226, 2013 www.atmos-meas-tech.net/6/1217/2013/